Discrete-time Adaptive Model Predictive Control based on Comparison Model

This paper proposes a discrete-time adaptive model predictive control (MPC) algorithm for a class of constrained linear time-invariant systems, which updates the estimation of system parameters on-line and produces the control input subject to the given input/state constraints. This method is based on a robust MPC algorithm using comparison model which enable us to estimate the prediction error bound of uncertain systems and an adaptive mechanism. First, a new parameter update method based on the moving horizon estimation is proposed, which allows us to predict the worst-case estimation error bound over prediction horizon. Second, we propose an adaptive MPC algorithm developed by combining the on-line parameter estimation with MPC method based on the modified comparison model. This method guarantees the feasibility and stability of closed-loop systems.